A GPU implementation for improved granular simulations with LAMMPS

E. MILLAN JUJTIUK; M.F. PICCOLI; C. GARCIA GARINO; E. BRINGA

Mendoza

Simposio; VI Latin American Symposium on High Performance Computing; 2013

UN Cuyo

Granular mechanics plays an important role in many branchesof science and engineering, from astrophysics applications in planetaryand interstellar dust clouds, to processing of industrial mixtures and powders. In this context, a granular simulation model with improved adhesion and friction, is implemented within the open source code LAMMPS (lammps.sandia.gov). The performance of this model is tested in both CPU and GPU (Graphics Processing Unit) clusters, comparing with performance for the LAMMPS implementation of another often usedinteraction model, the Lennard-Jones potential. Timing shows accelerations of 4-10x for GPUs versus CPUs, with good parallel scaling in a hybrid GPU-CPU cluster.